arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 1910.05084math.STstat.TH
收藏

将大范围流形拟合到噪声数据

Fitting a manifold of large reach to noisy data

Charles Fefferman, Sergei Ivanov, Matti Lassas, Hariharan Narayanan

摘要:

设${\mathcal M}\subset {\mathbb R}^n$是维数为$d$的$C^2$ -光滑紧致子流形.假设${\mathcal M}$的体积至多为$V$,并且${\mathcal M}$的到达范围(即,正常注入半径)大于$τ$。此外,设$μ$是${\mathcal M}$上的概率测度,其在${\mathcal M}$上的密度是严格正的Lipschitz光滑函数.令$x_j\in {\mathcal M}$、$j=1,2,\dots,N$是来自分布$μ$的$N$独立随机样本。同样,令$ξ_j$、$j=1,2,\dots, N$是来自具有协方差$σ^2I$的${\mathbb R}^n$中的高斯随机变量的独立随机样本,其中$σ$小于$d, V$和$τ$的某个指定函数。我们假设给定数据点$y_j=x_j+ξ_j,$ $j=1,2,\dots,N$,用测量噪声对${\mathcal M}$的随机点进行建模。我们开发了一种算法,从这些数据中产生,具有高概率,$d$维子流形${\mathcal M}_o\subset {\mathbb R}^n$的Hausdorff距离${\mathcal M}$小于$Cdσ^2/τ$,其范围大于$cτ/d^6$与通用常数$C,c > 0$。所需随机样本的数量$N$几乎线性地依赖于$n$,多项式地依赖于$σ^{-1}$,指数地依赖于$d$。

英文摘要:

Let ${\mathcal M}\subset {\mathbb R}^n$ be a $C^2$-smooth compact submanifold of dimension $d$. Assume that the volume of ${\mathcal M}$ is at most $V$ and the reach (i.e. the normal injectivity radius) of ${\mathcal M}$ is greater than $τ$. Moreover, let $μ$ be a probability measure on ${\mathcal M}$ whose density on ${\mathcal M}$ is a strictly positive Lipschitz-smooth function. Let $x_j\in {\mathcal M}$, $j=1,2,\dots,N$ be $N$ independent random samples from distribution $μ$. Also, let $ξ_j$, $j=1,2,\dots, N$ be independent random samples from a Gaussian random variable in ${\mathbb R}^n$ having covariance $σ^2I$, where $σ$ is less than a certain specified function of $d, V$ and $τ$. We assume that we are given the data points $y_j=x_j+ξ_j,$ $j=1,2,\dots,N$, modelling random points of ${\mathcal M}$ with measurement noise. We develop an algorithm which produces from these data, with high probability, a $d$ dimensional submanifold ${\mathcal M}_o\subset {\mathbb R}^n$ whose Hausdorff distance to ${\mathcal M}$ is less than $Cdσ^2/τ$ and whose reach is greater than $cτ/d^6$ with universal constants $C,c > 0$. The number $N$ of random samples required depends almost linearly on $n$, polynomially on $σ^{-1}$ and exponentially on $d$.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑